{"id":"67592d88-87fa-4709-931e-2137b2de3268","arxiv_id":"2505.10703","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"For the critical exponent b=b_c, IHQCD (alpha=1/2) admits only regular endpoint-type solutions at positive curvature and singular acceptable solutions at negative curvature, so the curvature-driven transition occurs at R=0.","lead":"This paper uses dynamical systems theory to classify the curved-space solutions of Improved Holographic QCD, a bottom-up holographic model of strongly coupled Yang-Mills. Its central finding is that, unlike other confining holographic models, IHQCD has no curvature-driven phase transition at positive curvature: the transition sits at zero curvature.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The non-existence of positive-curvature type I/II solutions is solid; the load-bearing gap is the companion claim of a phase transition at R=0, which rests on an unresolved log-fit extrapolation and a possibly inapplicable small-R expansion.","rationale":"The reader's verdict is CONDITIONAL with moderate confidence, citing both the sign-of-T derivation and the unresolved order of the R=0 transition. My stress-test agrees that the sign-of-T argument is actually more robust than the reader's weakest_assumption suggests, because T has constant sign along flows and the negative sign at large phi for each fixed C therefore forbids positive-curvature type I/II solutions globally. The real soft spot is the companion claim, repeated in the abstract, that a phase transition occurs at zero curvature. This is a separate assertion: it requires not just that Rc+=0, but that the free-energy densities from the two branches meet with some discontinuity. The paper's own Section 5.2.3 and 6 admit that the numerical accuracy is insufficient to resolve the order or even the existence of a genuine transition at R=0, and the small-curvature expansions used are imported from a setup with a regular IR fixed point, which is not the case here. The simplified UV potential (5.1) is also relevant here, since the free energy near R=0 is UV-sensitive through the conformal anomaly; however, I do not treat the UV simplification as the primary concern, because the solution-classification result is IR-dominated and robust. Since the reader already conditioned the verdict on precisely this unresolved transition, my concern does not move the verdict; it sharpens the reason. The concrete test I propose directly targets the missing analytic control: derive f(R) as R->0± from matched asymptotics rather than relying on log fits and numerical extrapolation.","tokens_in":46832,"tokens_out":11822,"duration_ms":130965,"concrete_test":"Derive the leading small-R behavior of the free-energy density f(R) for alpha=1/2 by matched asymptotic expansions: for type III use the endpoint expansion (A.1)-(A.10) as phi0->infinity, and for type I/II use the IR data (4.44)-(4.46) as C->-infinity. If both branches yield the same f(R) through all finite orders in R, or if the difference is flat at R=0, then no genuine zero-curvature phase transition exists and the abstract should be weakened. Cross-check numerically with phi_ini chosen so that log(phi_ini) >> |C| and with precision objective raised to 10^-40, including the next-order correction in (4.44); verify that the fitted R(C) and f(R) are stable under these changes. If a derivative discontinuity appears, the transition is real.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central non-existence claim for alpha=1/2 is well supported: Eq. (4.46) gives T negative for every fixed C at large phi, and because T=e^{-2A}R(zeta) has constant sign along flows, no type I/II solution can have positive boundary curvature. The load-bearing gap is the abstract's stronger statement that IHQCD has a phase transition at zero curvature. That requires comparing the free-energy density f(R) of type III solutions as R->0+ with type I/II solutions as R->0-. In Section 5.2.3 the authors state that for alpha=1/2 the numerics are 'not accurate enough' to determine whether any derivative of the free energy is discontinuous at R=0. Moreover, the small-R expansions (5.25)-(5.26) are taken from [19], where the zero-curvature endpoint is a regular IR fixed point; the authors themselves note it is unclear whether these expansions are valid here, where phi0->infinity and no such fixed point exists. The conclusion that Rc+=0 is also based on a log fit in Figure 10 that requires exponentially large phi0, with no independent analytic control. Thus the 'phase transition at zero curvature' is an extrapolation from the b>bc limit, not a demonstrated discontinuity, and the abstract overstates what the analysis establishes.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies Einstein-dilaton holographic models with critical IR asymptotics b = b_c and a power-law prefactor φ^α, focussing on the IHQCD value α = 1/2, for dual QFTs on constant-curvature backgrounds. The authors reformulate the large-φ asymptotic problem as an autonomous dynamical system, use center-manifold theory to classify the type I/II singular solutions and type III regular solutions, and then construct full numerical RG flows for a simplified UV potential. They derive an analytical expression for the slice curvature T in the IHQCD case, Eq. (4.46), whose sign is negative for all values of the integration constant, and use it to argue that no acceptable type I/II solutions exist for positive boundary curvature when α = 1/2. Numerically computed phase diagrams and free energies lead the authors to claim that for α > 1/2 a continuous phase transition occurs at a positive critical curvature R_{c+}, while for α < 1/2 and for α = 1/2 no finite-positive-curvature transition occurs; the abstract states that for IHQCD the phase transition occurs at zero curvature. The paper also presents the dynamical-system method as a novel tool for this class of holographic problems.","tokens_in":47106,"tokens_out":6702,"duration_ms":69908,"significance":"If the central non-existence claim is accepted, this is a significant result: the IHQCD-like critical asymptotics would be qualitatively different from all previously studied confining holographic models on curved backgrounds, since no curvature-driven transition occurs at positive curvature. The paper's main technical strengths are real: the center-manifold construction in Sections 3 and 4 is explicit and self-contained, the sign of T in Eq. (4.46) follows from an analytical solution of the reduced system rather than from numerical fitting, and the numerical procedure is described in enough detail in Appendix F to be reproducible. The exclusion of positive-curvature type I/II solutions for α = 1/2 is well supported. However, the abstract-level claim of a phase transition at zero curvature goes beyond what the paper actually demonstrates, because the free-energy comparison near R = 0 relies on an unresolved log-fit extrapolation and on small-R expansions whose applicability is uncertain. The paper is therefore a solid contribution with an overreaching summary statement.","major_comments":[{"comment":"The statement that for IHQCD 'the phase transition occurs at zero curvature' is not supported by the analysis presented. In Section 5.2.3 the authors explicitly state that for α = 1/2 the numerical data are 'not accurate enough' to determine whether any derivative of the free-energy density is discontinuous at R = 0, and Figures 16(c) and 17(c) show continuity of f and s only at the numerically accessible values of R. Moreover, the small-R expansions (5.25)-(5.26), taken from [19], were derived for flows that end at a regular IR fixed point in the zero-curvature limit, whereas here φ0 → ∞ and no such fixed point exists; the authors themselves flag this as unclear. The abstract and Section 6 should therefore be revised to state that the absence of a positive-curvature transition is established, while the behavior at R = 0 is a numerical extrapolation rather than a demonstrated phase transition.","section":"Abstract and Section 5.2.3"},{"comment":"The claim R_{c+} = 0 for α = 1/2 rests on a logarithmic extrapolation. The text explains that reaching smaller positive R requires exponentially larger φ0, making numerical computation exponentially slower, and the red curve in Figure 10 is a fit of the form R = 0.060/(4.8 + log φ0) with no independent analytic justification. Since the existence of type III solutions for all R > 0 is one half of the zero-curvature transition claim, this extrapolation should either be replaced by an analytic argument or be explicitly labeled as a numerical conjecture rather than part of the paper's main conclusions.","section":"Section 5.2.1 and Figure 10"},{"comment":"The numerical phase diagram is computed with the simplified UV potential (5.1), in which the marginally relevant Yang-Mills operator is replaced by a relevant operator of dimension Δ_- = 3/2. The paper asserts that this replacement does not change the qualitative features, but no comparison or argument is provided to show that the mapping from the IR constants to R, or the free-energy comparison across R = 0, is insensitive to this UV modification. Since the conclusions are framed as statements about IHQCD rather than about the toy potential, this gap should either be addressed explicitly or the claims should be restricted to the simplified class of potentials.","section":"Section 5, Eq. (5.1)"}],"minor_comments":[{"comment":"There is a typo in the sentence 'the numerical analysis is not accurate enough to indicates that the free energy density has some discontinuous derivative'; 'indicates' should be 'indicate', and the sentence should be rephrased for clarity.","section":"Section 5.2.3"},{"comment":"The word 'conicides' should be 'coincides' in the sentence describing the center manifold in d = 4.","section":"Section 3.4.2"},{"comment":"The sentence 'or more precisely, equation (5.24) below, which is derived from (5.24)' contains a self-reference error; the second occurrence of '(5.24)' should refer to the thermodynamic identity (5.18).","section":"Section 5.2.3"},{"comment":"The global phase diagram in Figure 14 would benefit from a legend clarifying the meaning of the dashed curves and the boundary between the orange and blue regions, since the caption currently describes these features only in the main text.","section":"Figure 14"},{"comment":"The renormalized free energy (5.21)-(5.22) depends on the finite counterterms A_ct, B_ct, C_ct, which are fixed to specific numerical values. The authors should state explicitly which reported conclusions are independent of this scheme choice, since a physical phase transition should not depend on the renormalization scheme.","section":"Section 5.2.3"},{"comment":"The phrase 'This is not crucial for the classification of the solutions and the IR dynamics' is stronger than what is demonstrated; the subsequent numerical results depend on the UV potential through the extraction of R and the free energy, so this point deserves a more qualified wording.","section":"Section 2.2"}],"recommendation":"major_revision","confidential_remarks":"The paper contains a strong analytical result — the negative sign of T in Eq. (4.46) — and the numerical work is careful. The main issue is that the abstract and Section 6 overstate the zero-curvature phase transition, which the authors themselves concede is not resolved at the level of numerical accuracy. I would ask for a revision that either provides an analytic argument for the R → 0 behavior or explicitly presents the zero-curvature conclusion as a conjecture. The reliance on companion papers for the holographic dictionary and free-energy formulas is acceptable for a JHEP submission, but the authors should ensure that the distinctions between established results, numerical evidence, and extrapolation are transparent."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the IR analysis at the critical exponent b = bc. The center-manifold treatment is a real technical advance, and it pays off: for alpha = 1/2 the sign of the slice curvature T in eq. (4.46) is negative for every integration constant, so no positive-curvature type I/II solutions exist. That is an analytic, parameter-free statement, and I trust it. The supporting machinery -- three-dimensional autonomous system, the Riccati reduction, the exact d=4 Lambert-function solution in Appendix D -- is careful and internally consistent. The paper also does honest numerical checks (Appendix F) and flags its own limitations, which is more than most papers do.\n\nThe soft spots are real but narrower than the abstract suggests. The claim that IHQCD has a phase transition at R=0 is an extrapolation. In Section 5.2.3 the authors admit the numerics are not accurate enough to determine whether any derivative of the free energy is discontinuous at R=0. The small-R expansions (5.25)-(5.26) come from a setup with a regular IR fixed point, and the authors themselves note it is unclear whether they apply when phi0 -> infinity. So the abstract's \"phase transition occurs at zero curvature\" is too strong; what is established is that no positive-curvature type I/II solutions exist, and that the type III branch extends down to R=0. The order of the transition, or whether there is one, remains open.\n\nThe simplified UV potential (5.1) is a legitimate simplification but a caveat: it replaces the marginally relevant YM operator by a relevant one, and the qualitative phase structure is assumed to survive. That assumption is plausible but not proven. No code or data are provided, though the numerical method is described in enough detail to reproduce.\n\nWho is this for? People working on holographic QCD, curved-space holography, and dynamical systems in gravity. The central non-existence result should be cited. It deserves a serious referee: the formal part is strong, and the physics claim, once softened, is worth publishing. I would send it to review, but I would ask the authors to either prove the discontinuity or explicitly reframe the abstract to say the transition boundary is at R=0 without claiming a proven phase transition.","headline":"The analytic sign-of-curvature result for IHQCD is solid and new, but the abstract's claim of a phase transition at zero curvature outruns the numerics, which only show continuity of the free energy, not a discontinuity.","tokens_in":47665,"tokens_out":2074,"would_cite":true,"duration_ms":21948,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"For the holographic QCD model IHQCD, the quantum phase transition driven by boundary curvature occurs at zero curvature, not at finite positive curvature.","keywords":["Improved Holographic QCD","Einstein-dilaton gravity","curvature phase transition","dynamical systems","center manifold","confining gauge theory","de Sitter space","holographic QCD"],"falsifier":"Integrate the full second-order equation (5.2) with the IHQCD potential (5.1) at $\\alpha=1/2$, starting from the type I/II infrared asymptotics (4.44) and scanning negative $C_{1/2}$; if any solution reaches the ultraviolet fixed point with dimensionless curvature $R>0$, the central claim is false. Equivalently, compute the full nonlinear slice curvature $T(\\varphi)$ along the type I/II branch and look for a sign change away from the asymptotic region.","tokens_in":46578,"feed_emoji":"⚛️","tokens_out":7223,"duration_ms":65825,"temperature":0.7,"pith_summary":"This paper asks what happens when Improved Holographic QCD (IHQCD), a bottom-up holographic model of large-$N_c$ Yang-Mills theory, is placed on a spacetime of constant curvature. Earlier work on confining Einstein-dilaton models with exponential infrared potentials found a quantum phase transition at finite positive curvature. The paper shows that at the critical exponent $b=b_c$, where IHQCD lives, the sign of the slice curvature for the singular type I/II solutions is negative for the physical value $\\alpha=1/2$, so no such solutions exist for positive curvature. Consequently the curvature-driven phase transition in IHQCD occurs at zero curvature, not at a finite positive value; for $\\alpha>1/2$ a finite positive critical curvature reappears, while for $\\alpha<1/2$ only negative curvature admits type I/II solutions. A sympathetic reader would care because this distinguishes the most phenomenologically successful confining holographic model from every other model in its class.","feed_headline":"Curving space no longer triggers a phase transition in holographic QCD","feed_subtitle":"The IHQCD model of Yang-Mills theory loses its finite-curvature transition, unlike other confining holographic models.","key_machinery":"The load-bearing object is the autonomous dynamical system in the variables $(\\tilde W,\\tilde S)$, and the auxiliary variable $Z=-\\alpha/\\varphi$ for $\\alpha\\neq 0$, obtained from the superpotential and scalar velocity; its non-hyperbolic fixed point represents the type I/II asymptotic solutions. Around this fixed point the paper applies center manifold theory, the nonlinear generalization of the zero-eigenvalue eigenspace, expanding the unique center manifold to second order and pulling the flow back onto it to reduce the system to a Riccati equation for $U$. The roots $1$ and $2\\alpha$ collide at $\\alpha=1/2$, turning the power-law correction into a logarithmic one, and the sign of the resulting slice curvature $T$ in equation (4.46) carries the physical conclusion: $T<0$ for all $C_{1/2}$, excluding positive-curvature type I/II solutions and placing the phase transition at $R=0$.","core_discovery":"The paper's central claim is that for Einstein-dilaton theories with infrared potential $V(\\varphi)\\sim -V_\\infty \\varphi^\\alpha e^{2b_c\\varphi}$, where $b_c=1/\\sqrt{2(d-1)}$, the curvature dependence of the ground states changes at $\\alpha=1/2$. Using a center-manifold reduction of the autonomous dynamical system describing the large-$\\varphi$ asymptotics, the authors compute the slice curvature $T(\\varphi)$ for the type I/II solutions. For $\\alpha=1/2$ the subleading correction is logarithmic and $T$ is negative for every value of the integration constant $C_{1/2}$; hence type I/II solutions exist only for $R\\le 0$, and all positive-curvature regular ground states are type III solutions ending at finite field value. The numerical phase diagram then shows that the transition between type III and type I/II branches occurs at $R=0$ for $\\alpha=1/2$, with continuous free energy density and entropy there, so the curvature-driven phase transition of IHQCD is at zero curvature.","pith_inferences":["The paper leaves implicit that if the zero-curvature transition is real for the model, then on a large-radius sphere IHQCD stays in the type III phase all the way to the flat limit; computing the spectrum on $S^4$ as $R\\to 0^+$ would test whether observables vary smoothly there.","Because both the type III and type I/II phases are gapped and the paper identifies no order parameter separating them, a conservative reading is that boundary curvature is not an order parameter for IHQCD, weakening the notion of a curvature-driven deconfinement transition in this model.","The logarithmic correction at $\\alpha=1/2$ makes numerical access to very small positive $R$ exponentially expensive, so settling the order of the $R=0$ transition will likely require an analytic small-$R$ expansion of the type III branch.","For $\\alpha<1/2$ the phase diagram is asymmetric between positive and negative curvature, suggesting that the corresponding dual field theories on positively and negatively curved slices flow to qualitatively different infrared states; combining this with the negative-curvature classification for non-critical $b$ would give a unified picture."],"forward_implications":["For IHQCD ($\\alpha=1/2$) every regular ground state on a positively curved slice is a type III solution ending at finite field value; singular type I/II solutions are confined to $R\\le 0$.","The curvature-driven quantum phase transition, present at finite positive $R$ for $\\alpha>1/2$, is pushed to $R=0$ exactly at $\\alpha=1/2$, and it remains continuous (second or higher order) there.","In the $R$--$\\alpha$ phase diagram, type III and type I/II branches never overlap, and $\\alpha=1/2$ is a double bifurcation point where both $R_{c+}$ and $R_{c-}$ approach zero.","IHQCD is thus the first confining holographic model studied in this framework with no finite-curvature phase transition, even though it still confines in flat space and has a finite-temperature deconfinement transition.","The center-manifold method applies beyond the infrared: with an auxiliary constrained variable it can handle full RG flows and multi-field theories, so the same classification can be repeated for richer holographic models."],"supporting_citations":[{"why":"Supplies the previous classification of positive-curvature solutions for $b>b_c$ and the finite-curvature phase transition that this paper extends to the critical case $b=b_c$.","marker":"[7]"},{"why":"Provides the negative-curvature classification for $b>b_c$, including type I/II solutions, bounces, and the sign-of-$T$ analysis used for comparison.","marker":"[16]"},{"why":"Defines IHQCD and fixes the infrared potential asymptotics $V\\sim -V_\\infty \\varphi^\\alpha e^{2b_c\\varphi}$ with $\\alpha=1/2$ and the Regge glueball spectrum.","marker":"[5]"},{"why":"Introduces the $W,S,T$ variables, the dimensionless curvature parameter $R$, and the type III endpoint asymptotics on which the numerical construction relies.","marker":"[17]"},{"why":"Gives the holographic free-energy and entanglement-entropy formulas, including finite counterterms used to extract the phase-transition order.","marker":"[18]"},{"why":"Provides small-curvature expansions for $B(R)$ and $C(R)$ used in the free-energy fits.","marker":"[19]"},{"why":"Supplies the center manifold theorem used to reduce the non-hyperbolic fixed point and derive the asymptotic solutions.","marker":"[23]"},{"why":"Supplies the bifurcation-theory background for transcritical bifurcations and the Hartman-Grobman theorem.","marker":"[24]"},{"why":"Establishes the finite-temperature deconfinement transition in IHQCD, the counterpart to the curvature transition studied here.","marker":"[28]"}],"fun_headline_variants":["Curved space can't trigger holographic QCD phase transition","Holographic QCD phase transition only at zero curvature","IHQCD: curvature-induced phase transition vanishes","Holographic QCD's transition pinned to flat space"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The exclusion of positive-curvature type I/II solutions in IHQCD rests on assuming that the negative sign of the slice curvature obtained from the leading large-$\\varphi$ asymptotic expansion persists for the full nonlinear solutions away from the asymptotic regime, and that replacing the marginally relevant Yang-Mills operator by a relevant ultraviolet operator does not alter the qualitative phase structure.","fun_headline_variants_meta":{"raw":{"variants":["Curved space can't trigger holographic QCD phase transition","Holographic QCD phase transition only at zero curvature","IHQCD: curvature-induced phase transition vanishes","Holographic QCD's transition pinned to flat space"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001194,"raw_usage":{"total_tokens":4902,"prompt_tokens":900,"completion_tokens":4002,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":516,"completion_tokens_details":{"reasoning_tokens":3936}},"tokens_in":516,"tokens_out":4002,"duration_ms":30168,"temperature":1.0,"reasoning_tokens":3936,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:04:39.185747+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Integrate the full second-order equation (5.2) with the IHQCD potential (5.1) at $\\alpha=1/2$, starting from the type I/II infrared asymptotics (4.44) and scanning negative $C_{1/2}$; if any solution reaches the ultraviolet fixed point with dimensionless curvature $R>0$, the central claim is false. Equivalently, compute the full nonlinear slice curvature $T(\\varphi)$ along the type I/II branch and look for a sign change away from the asymptotic region.","supporting_citations":[{"cited_title":"Applications of center Manifold Theory,","cited_arxiv_id":null,"evidence_quote":"Supplies the center manifold theorem used to reduce the non-hyperbolic fixed point and derive the asymptotic solutions."},{"cited_title":"Guckenheimer and P","cited_arxiv_id":null,"evidence_quote":"Supplies the bifurcation-theory background for transcritical bifurcations and the Hartman-Grobman theorem."}],"review_version":1}